Doctor of Philosophy
School of Civil, Mining and Environmental Engineering
Fyfe, Julian, Characterisation, monitoring and dynamic modelling of a two-stage stabilisation pond system supporting reuse and recycling of dairy shed effluent, Doctor of Philosophy thesis, School of Civil, Mining and Environmental Engineering, University of Wollongong, 2013. http://ro.uow.edu.au/theses/4263
Best practice handling of dairy shed effluent (DSE) in Australia and New Zealand currently uses two-stage stabilisation pond systems, comprising anaerobic and facultative ponds in series. One of the purported benefits of two-pond systems is the ability to contain and reuse the nutrients contained in the manure deposited at the dairy. Estimates of recoverable nutrient loads, however, are typically based upon coarse partitioning factors, and understanding remains limited regarding the complex array of physical, chemical and biological processes occurring in DSE ponds and their effects on nutrient forms and quantities. There is also very little information available on the impacts of placing pond systems within a partially closed effluent recycling loop in which effluent is recycled back to the dairy as flush water or reused via irrigation. This thesis aims to address the shortage of data related to the behaviour and performance of Australian (and New Zealand) DSE pond systems supporting recycling, which are distinct from systems elsewhere for their relatively small manure loads and more dilute influent. It does so through a detailed field-based study of a typical best practice system which treats effluent from a commercial pasture-based dairy farm with a milking herd of 300 cows. The field data are used to establish a foundation for dynamic modelling of these systems, leveraging knowledge from the more developed modelling fields of urban wastewater stabilisation ponds and activated sludge systems.
High resolution water quality monitoring, incorporating real time measurement at various locations in the system and seasonal profiling of the water column of each pond, produced a unique insight into temporal (including diurnal, seasonal and longer term) and spatial variation in the underlying conditions that both inform and respond to the natural treatment processes occurring in DSE ponds. Important observations included spatial uniformity of water quality in the anaerobic pond, biochemical and thermal stratification in the facultative pond, and persistent salt accumulation in both ponds.
A water balance analysis using wastewater flow and storage data and involving calibration and validation of detailed evaporation and seepage models generated high resolution data on the hydrology of pond system. An in-depth analysis of pond hydraulics and hydrodynamics used hydrological, meteorological, sludge accumulation, water quality and in-pond Lagrangian flow data to characterise the hydraulic regime of each pond. Some key insights were the loss of active treatment volume to accumulating sludge (at a rate of 0.0043 m3 kg-1 total solids added or 0.88 m3 cow-1 y- 1), the role of rising biogas bubbles in producing well-mixed conditions in the anaerobic pond, the low frequency of short-circuiting (5% or less) and the potential presence of a hydraulic dead zone in the anaerobic pond, and the complexity of the hydraulics in the facultative pond despite very slow advective flow.
Data generated through an intensive flow-weighted wastewater sampling program was then used to characterise the influent to and effluents from the pond system and the dominant treatment and nutrient cycling processes in each stage of the system. This enabled effluent variability, biodegradability and constituent fractions and loading, as well as other aspects of pond function to be studied in more detail than had previously been possible, particularly the effects of sludge accumulation and struvite precipitation (related to salt accumulation caused by recycling). A mass balance analysis was also developed to quantify partitioning and removal of wastewater constituents and investigate actual and potential nutrient recovery rates. Contrary to conventional assumptions, nutrient partitioning to the sludge in the anaerobic pond was found to be lower than expected at just 18% for phosphorus (P) and 21% for nitrogen (N), and overall nutrient removal was actually higher in the facultative pond. Low effluent nutrient utilisation rates (47% N and 50% P) and problematic accumulation of potassium (K) helped to identify a new management strategy for pond systems supporting effluent recycling whereby the pond system is used to harvest stormwater to dilute recycled effluent and improve nutrient recovery rates.
The outputs from the monitoring and characterisation work represent a comprehensive set of qualitative and quantitative data unprecedented in Australia and internationally and provided the basis to developing a dynamic biokinetic model of the primary anaerobic pond using the Activated Sludge Anaerobic Digestion model (ASDM) in the BioWin simulation environment. The model was formulated with a simplified yet dynamic hydraulic regime, full hydrological accounting, and a comprehensive suite of process sub-models coupled with state variables to represent detailed fractionations of organic substrate, nutrients and salts. Model calibration achieved reasonable agreement between predicted and observed concentrations for total and filterable chemical oxygen demand (11% and 16% mean absolute percentage error, respectively), total suspended solids (11%), total volatile suspended solids (9%), total N (14%), ammonia N (11%), total P (10%), orthophosphate (12%), calcium (7%), magnesium (7%) and K (7%). It also exhibited sensible responses to the shock induced by desludging and produced predictions of biogas production which were used to estimate greenhouse gas emissions from the pond.